sociolinguistic survey of northern pakistan volume 4

8/14/2019 Sociolinguistic Survey of Northern Pakistan Volume 4

1/194

SOCIOLINGUISTIC SURVEY OF

NORTHERN PAKISTAN

VOLUME 4

PASHTO, WANECI, ORMURI


2/194

Sociolinguistic Survey of Northern Pakistan

Volume 1 Languages of Kohistan

Volume 2 Languages of Northern Areas

Volume 3 Hindko and Gujari

Volume 4 Pashto, Waneci, Ormuri

Volume 5 Languages of Chitral

Series Editor

Clare F. OLeary, Ph.D.


3/194

Sociolinguistic Survey

of

Northern PakistanVolume 4

Pashto

Waneci

Ormuri

Daniel G. Hallberg

National Institute of

Pakistani StudiesQuaid-i-Azam University

Summer Institute

ofLinguistics


4/194

Copyright 1992 NIPS and SIL

Published by National Institute of Pakistan Studies,Quaid-i-Azam University, Islamabad, Pakistan

and

Summer Institute of Linguistics, West Eurasia Office

Horsleys Green, High Wycombe, BUCKS

HP14 3XL United Kingdom

First published 1992

Reprinted 2004

ISBN 969-8023-14-3

Price, this volume: Rs.300/-

Price, 5-volume set: Rs.1500/-

To obtain copies of these volumes within Pakistan, contact:

National Institute of Pakistan Studies

Quaid-i-Azam University, Islamabad, Pakistan

Phone: 92-51-2230791

Fax: 92-51-2230960

To obtain copies of these volumes outside of Pakistan, contact:

International Academic Bookstore

7500 West Camp Wisdom Road

Dallas, TX 75236, USA

Phone: 1-972-708-7404

Fax: 1-972-708-7433

Internet: http://www.sil.org

Email: [email protected]

REFORMATTING FOR REPRINT BY R.CANDLIN.


5/194

CONTENTS

Preface.............................................................................................................viiMaps................................................................................................................. ix

Introduction Clare F. OLeary...................................................................xiAcknowledgments ........................................................................................ xviiChapter 1 Pashto: A Sociolinguistic and Dialect Study ............................... 1

1. Framework and Purpose ...................................................................... 12. General Background............................................................................ 2

2.1 Pashto Language Classification.................................................... 22.2 The Geographic Boundaries of Pashtoon Territory ...................... 22.3 Population Estimates .................................................................... 52.4 Education in Pakistan ................................................................... 6

2.5 Comments on the Pashto Alphabet as Expressed in the Literature73. Dialects of Pashto ................................................................................ 93.1 Pashto Dialect Groupings as Expressed in Existing Literature..... 93.2 Word List Findings..................................................................... 143.3 Comprehension of Recorded Speech.......................................... 203.4 Dialect Groupings Based on Native Speaker Opinion................ 223.5 Pashto Dialect Groupings Summary and Conclusions........... 25

4. Bilingualism / Second Language Proficiency.................................... 274.1 Urdu ........................................................................................... 274.2 English ....................................................................................... 29

5. Language Use and Attitudes.............................................................. 305.1 Comments on Language Use and Attitudes as Expressed in

Existing Literature.................................................................... 305.2 Language Use as Reported by Interview Subjects...................... 355.3 Language Attitudes as Expressed by Interview Subjects............ 39

6. Summary and Conclusions ................................................................ 43Chapter 2 Wan4eci .......................................................................................... 45

Introduction and Purpose....................................................................... 451. Discussion of Existing Literature on Wan4eci..................................... 45

1.1 The Language............................................................................. 451.2 The WaNeci People .................................................................... 47

2. Word List Findings............................................................................ 503. Discussion of Waneci (Tarino) Interview Data.................................. 504. Summary and Conclusions ................................................................ 525. Further Study..................................................................................... 52

Chapter 3 Ormur4i ......................................................................................... 53Introduction and Purpose....................................................................... 531. Setting................................................................................................ 532. Dialects of Ormur4i ............................................................................. 55

3. Comparison of Ormur4i and Pashto .................................................... 564. Second Language Proficiency (Bilingualism).................................... 595. Language Use and Attitudes.............................................................. 606. Language Vitality .............................................................................. 617. Summary and Conclusions ................................................................ 63


6/194

Appendix A Methodologies ........................................................................... 67A.1 Procedure for Counting Lexical Similarity ..................................... 67A.2 Recorded Text Testing.................................................................... 75

RTTs in Second Language Testing .................................................. 77Appendix B Word Lists ................................................................................. 79Appendix C Texts......................................................................................... 147

C.1 Peshawar Pashto, Peshawar......................................................... 147C.2 Peshawar Pashto, Peshawar......................................................... 152C.3 Quetta Pashto, Quetta.................................................................. 159C.4 Ormuri (Kaniguram, South Waziristan)...................................... 165

Appendix D Questionnaire .......................................................................... 171References..................................................................................................... 175


7/194

PREFACE

The northern area of Pakistan occupies a unique position onthe cultural and historical map of the world. Its cultural diversityand ethnic richness make it one of the most fascinating areas forresearchers and scholars. It is, however, its multi-lingualcharacter that concerns the present study.

These five volumes of the Sociolinguistic Survey of Northern Pakistan are devoted to the study of its multi-lingualfeatures. It is slightly more ambitious than the usual studies of

this nature: it attempts to study the various languages anddialects of this area from a synchronic descriptive approach withregard to the issue of language versus dialect. In order to verifythe diversity and similarity within these languages and dialects,linguistic and sociolinguistic data has been used to throw somelight on the relative levels of diversity within and between theidentified varieties. This has been done particularly in the casesof Gujari with Hazara Hindko, Indus and Swat varieties of

Kohistani and Shina with its linguistic neighbours.

At a macro level, this work is definitely an improvementover Griersons Linguistic Survey of India and the subsequentstudies by various scholars. However, though ambitious in scope,the study does not claim to be exhaustive and comprehensive inevery respect. The study also discusses the impact of externallinguistic families on the linguistic evolution of this area. Theunmistakable imprint of Tibeto-Burman languages, the Iranian

languages, the Indo-European family and the Indo-Aryan familytestify to the fact that the northern areas of Pakistan serve as abridge between South Asia, Central Asia, China, and Iran.

Another dimension has also been added to the study of somany languages and dialects in close proximity: degree of proficiency in the neighbouring languages. This has been donethrough interviews, questionnaires, tests, and observations. The patterns associated with the proficiency of the neighbouringlanguages and the national language, Urdu, are treated in termsof inter-ethnic contacts, the regional dominance of certainlinguistic groups, and the impact of education and media. It is


8/194

quite visible that the old generation of these linguistic groups didtry to preserve the originality of their culture and civilization. Butcommunication links and the availability of modern techniques

and instruments have their own impact upon the people of theseareas. The new generation of these areas, showing a trendtowards advancement and modernization, may in the long run beaffected, and the preservation of centuries old culture andcivilizations can become a difficult task.

It is hoped that this survey will inspire some studies of thisunique multi-linguistic region of the world. The scholars deservecongratulations for this painstaking work, which could not have

been completed without requisite enthusiasm, expertise and skill.This study, of course, will open new avenues for futureresearchers. The important point to be kept in mind for futureresearchers is, however, to find ways and means of preservingthis centuries old culture and civilization.

Work of such a magnitude is not possible withoutcooperation and devotion on the part of scholars and experts inthis field. The National Institute of Pakistan Studies, Quaid-i-

Azam University, Islamabad acknowledges with gratitude theassistance and cooperation of many who helped the team toconduct this survey. The Institute acknowledges the commitmentof the Summer Institute of Linguistics (the co-sponsors of thisproject), the Ministry of Culture Government of Pakistan, andthe National Institute of Folk Heritage for providing all sorts ofhelp to complete this study. The Institute feels honored forhaving such association with these institutions as well as the

scholars of repute who devoted their precious time and expertisein preparing this important study.

The National Institute of Pakistan Studies will feel happy inextending maximum cooperation to the scholars interested inexploring further studies in the field.

Dr. Ghulam Hyder SindhiDirector

National Institute of Pakistan StudiesQuaid-i-Azam UniversityIslamabad, Pakistan


9/194

MAPS


10/194


11/194

INTRODUCTION

Northern Pakistan is a land of geographic and ethnicdiversity, one of the most multilingual places on the face of theearth. Spectacular mountain ranges and mighty rivers segmentthe area, providing natural barriers which often serve asisoglosses separating linguistic varieties. Centuries of peoplemovements across this crossroad of South and Central Asia have

left a complex pattern of languages and dialects, fertile groundfor sociolinguistic investigation.

Twenty-five named languages from within northernPakistan are dealt with in the volumes of the SociolinguisticSurvey of Northern Pakistan. Most languages of the region havebeen classified as part of the large Indo-Aryan (or Indic) family.Two of these have been called members of the Central Groupaccording to the scheme established in Griersons Linguistic

Survey of India: Gujari, subgrouped with other Rajasthanilanguages, and Domaaki, not even mentioned by Grierson, butclassified as Central by Fussman (1972) and Buddruss (1985). Athird named language, Hindko, was originally included withinthe Northwestern Group of Indo-Aryan, among those varietieswhich were given the label Lahnda (LSI VIII.1). The variousforms called Hindko have been particularly difficult to classify(Shackle 1979, 1980), showing a wide geographic range, much

linguistic divergence, and some convergence with Panjabi, whichhas been classified in the Central Group.

The largest number of Indo-Aryan languages dealt with inthese volumes belong to the Northwestern Group, Dardic branch:Shina, and its historical relations, Phalura and Ushojo; IndusKohistani, and its smaller neighbors, Chilisso, Gowro, and, presumably, Bateri (which has not been classified); the SwatKohistani varieties, Kalami and Torwali; the Chitral group ofKhowar and Kalasha; and the Kunar group, including Dameliand Gawar-bati. The Nuristani branch accounts for somelanguages spoken on the northwestern frontier; within Pakistanthat group is represented by Eastern Kativiri andKamviri/Shekhani. This classification outline for members of the


12/194

Introductionxii

Dardic and Nuristani branches is based on several scholarlycontributions (Fussman 1972, Masica 1991, Morgenstierne1932), but primarily follows Strand (1973).

There are also members of the larger Iranian family(classification following Payne 1987). Some come from theSoutheastern Iranian group, the major example being Pashto, butalso including the more divergent WaNeci. Others are from theSoutheastern Iranian Pamir subgroup: Wakhi and Yidgha.OrmuRi has been classified as a Northwestern Iranian languagebut shows the influence of being surrounded by Pashto.

Finally, a few linguistic relics remain from outside thelarger Indo-European family, notably the westernmost Tibeto-Burman language, Balti, and the isolate, Burushaski.

The distinction between language and dialect is always afuzzy one, but particularly so in this part of the world. Scholarshave long acknowledged the immense dialect continuum whichcharacterizes the South Asian region, particularly among theIndo-Aryan varieties. The difficulties in drawing language

distinctions are compounded by the terminological confusionfound when local speakers use identical names to label their verydifferent spoken varieties (e.g., Kohistani) or apply the name of alarger and more prestigious language to cover a very wide rangeof speech forms (e.g., Panjabi).

Rather than focussing on linguistic classification or on thehistorical relationships between languages, the SociolinguisticSurvey of Northern Pakistan has taken a synchronic descriptiveapproach to this issue of language versus dialect. Linguistic andsociolinguistic data to verify the diversity and similarity withinthe varieties have been collected for all twenty-five namedlanguages. These data include a consistent 210-item word listfrom several locations within a language group. In addition, oraltexts have been recorded and transcribed from many locations;often these texts have been used to assess the intelligibility ofspoken forms among speakers of divergent dialectal varieties.

Word list comparisons have been made across named languagesin some cases (e.g., Gujari with Hazara Hindko, Indus and Swatvarieties of Kohistani, Shina with its linguistic neighbors), to


13/194

Introduction xiii

give some perspective on the relative levels of diversity withinand between named varieties. These comparisons of linguisticdata are balanced by information gathered through interviews

and orally-administered questionnaires regarding ethnicidentification, dialect group contacts, and perceived linguisticsimilarity and difference. Although few sharp boundaries areevident, groupings of relatively similar varieties can bedemonstrated according to the criteria of lexical similarity,indications of intelligibility, patterns of within-group contact, anddialect perceptions of the speakers themselves.

The investigation of local language names has provided a

perspective on the linguistic identification of its speakers. Whereit is possible to use the locally preferred name without ambiguity,those local names have been chosen to designate the linguisticvarieties described in these volumes. Where further clarificationis necessary, language names have included regionaldesignations or have incorporated the labels given by previousscholars even though they were not found to be used by thespeakers themselves.

In addition to questions of diversity within languages, thereare higher levels of sociolinguistic variation which are evident inthe prevalence of multilingualism throughout the area. Ingeneral, it seems that members of most language groups innorthern Pakistan exhibit pragmatic attitudes toward adoption oflanguages of wider communication. With so many languages inclose proximity, it is commonplace for persons to acquire one ormore of their neighboring languages to some degree of proficiency. Some studies included tests of proficiency in thenational language, Urdu, or in a regional language of widercommunication such as Pashto or Hindko. Other reports haveinvestigated reported proficiency and use of other languagesthrough interviews, orally-administered questionnaires, andobservation. The patterns associated with the use of otherlanguages are related to such social phenomena as inter-ethniccontacts, the regional dominance of certain groups, and the

promotion of Urdu through education and the media. A fewlanguage groups indicate signs of declining linguistic vitality andthe preference for more dominant neighboring languages among


14/194

Introductionxiv

the younger generations within those groups (e.g., Domaaki,Chilisso, Gowro, Yidgha). But, for the present, most of the ethniclanguages of northern Pakistan are well-maintained by their

mother-tongue speakers as the most frequently used andapparently valued means of communication.

A major contribution of the Sociolinguistic Survey ofNorthern Pakistan is the collection of the standard 210-item wordlist; combining the lists from all twenty-five languages yields asum of 127 regional speech forms represented. The phoneticallytranscribed lists for the reports covered in each volume are presented in the relevant appendices. Story texts for the

languages represented are presented as well, with a rough word-for-word gloss and a free translation. In total, there are forty-ninetranscribed texts in these volumes. This fieldwork has notundergone thorough grammatical and phonological analysis; it isincluded to support the conclusions presented in each report andas data for future scholarship.

In terms of methodology, this research makes a contribution

as well. A multipronged approach was utilized in each study,combining some or all of the following: participant observation,interviews and orally-administered questionnaires, testing ofsecond language proficiency, testing of comprehension of relatedvarieties, and the comparison of word lists by a standardizedmethod measuring phonetic similarity. Overall, the data showgreat internal consistency, with many types of self-reports fromquestionnaires and interviews corresponding well with moreobjective measures such as test results and lexical similaritycounts.

Each report reflects a slightly different focus. Someemphasize interdialectal variation and intelligibility (e.g., Balti,Burushaski, Pashto, Shina, Wakhi); others include this focus, butconcentrate more than the rest on assessing the proficiency anduse of other languages (e.g., the reports on the languages ofIndus and Swat Kohistan, Gujari, Hindko). The high

concentration of languages in the Chitral region makemultilingualism and ethnolinguistic vitality a primary concern inthat volume. Issues of declining vitality are of critical concern for


15/194

Introduction xv

Domaaki. One language included in this research has not been previously described or reported: Ushojo, a variant of Shinalocated in the Chail Valley of Swat District.

It has been a privilege to work with representatives of eachof these ethnolinguistic groups in carrying out this surveyresearch. These volumes are offered in the hope that they will provide a holistic overview of the sociolinguistic situation innorthern Pakistan and will stimulate further such work in the

years to come.

Clare F. OLearySeries Editor

References

Buddruss, Georg. 1985. Linguistic research in Gilgit and Hunza: Some resultsand perspectives.Journal of Central Asia 8:1.27-37.

Fussman, Gerard. 1972. Atlas linguistique des parlers dardes et kafirs. 2 vols.Paris: Ecole Francaise dExtreme Orient.

Grierson, George A. 1903-28. Linguistic survey of India, vols. I-XI. Calcutta.[reprinted 1968, Delhi: Motilal Banarsidass.]

Masica, Colin P. 1991. The Indo-Aryan languages. Cambridge: CambridgeUniversity Press.

Morgenstierne, Georg. 1932. Report on a linguistic mission to north-westernIndia. Oslo: Instituttet for Sammenlignende Kulturforskning.

Payne, J. R. 1987. Iranian languages. The worlds major languages, ed. byBernard Comrie, 514-522. Oxford: Oxford University Press.

Shackle, C. 1979. Problems of classification in Pakistan Panjab. Transactionsof the Philological Society 1979, 191-210.

. 1980. Hindko in Kohat and Peshawar. Bulletin of the School ofOriental and African Studies 43:3.482-510.

Strand, Richard F. 1973. Notes on the Nuristani and Dardic languages.Journalof the American Oriental Society 93:3.297-305.


16/194


17/194

xvii

ACKNOWLEDGMENTS

Acknowledgments for the Series

The Sociolinguistic Survey of Northern Pakistan was carriedout by researchers from the Summer Institute of Linguistics inconjunction with Pakistani scholars and research assistants. Thesurvey began in 1986 under the auspices of the Ministry ofCulture, through the National Institute of Folk Heritage, Lok

Virsa. Arrangements for the project were greatly facilitated bythe kind help of His Excellency Ambassador JamsheedK. A. Marker, who served at the Embassy of Pakistan,Washington, D. C. at the time the project was begun. Gratefulappreciation is also due to Dr. G. A. K. Niazi, former EducationAttache, Embassy of Pakistan in Washington, D. C. The Directorof Research for Lok Virsa, Dr. Adam Nayyar, gave oversight anddirection, particularly in the early stages of the project; hisassistance is appreciated. Dr. Naveed-i-Rahat, chairman of theDepartment of Anthropology at Quaid-i-Azam University, wassupportive of the research goals of this survey and gave adviceconcerning the project. Through her help, several graduatestudents from the department became involved in the project asresearch assistants in the study of the ethnolinguistic minoritiesin northern Pakistan.

The National Institute of Pakistan Studies of Quaid-i-Azam

University has a continuing commitment to the promotion ofquality research; without the assistance of NIPS, these volumeswould not have been completed. Special thanks are due toProfessor Fateh Mohammad Malik, who made initialarrangements for the publication of these studies, and toDr. Ghulam Haider Sindhi, who has worked with the researchteam to see the publication through to completion.

We have benefited from interaction with scholars interested

in linguistic studies in Pakistan. In particular, we have receivedinvaluable input from Dr. Aurangzeb Shah, former chairman ofthe Department of English at the University of Peshawar. Hegave assistance in many practical ways and also introduced us toyoung scholars who participated in the research project. The


18/194

Acknowledgmentsxviii

support of Dr. Sahibzada M. Riaz, current chairman of thedepartment, is appreciated as well.

Dr. Calvin Rensch served as the project director for theSociolinguistic Survey of Northern Pakistan. Ongoing leadershipwas also provided by Mr. Donald Gregson and Mr. DanielHallberg. Mr. Donald Johnson gave timely assistance in theinitial stages of the project. The preparation of the finalmanuscripts was coordinated by Miss Carla Radloff, who wasably assisted by Mrs. Sandra Decker. Mr. Peter Mielke and

Ms. Beverly Garland each provided critically needed help in the preparation and checking of these volumes. Special thanks aredue to Mr. Kendall Decker for map production and cover design.It has been a privilege to work with such a team of colleagues.

Clare F. OLeary

Acknowledgments for this volume

This sociolinguistic and dialect study of Pashto was carriedout over a period of about two years, from 1988 through late1990. The main research assistants involved in the project wereAmjad Afridi, Imtiaz Afridi, and Rahmanullah Shinwari. Thesemen acted as excellent advisors and, in general, greatlyfacilitated the entire data gathering process. In many cases they

dealt with conditions of substantial hardship in order to gatherwhat was needed. Much thanks must also be given to the manyindividuals throughout the Pashtoon areas of Pakistan who werekind enough to give of their time. In particular, thanks should begiven to Mr. Darwesh, who at the time I was in Quetta, offeredvaluable information concerning the Pashto of that area. Specialthanks should also be given to all those who demonstrated greatkindness to me through their generous hospitality.

Daniel G. Hallberg

July 1992


19/194

CHAPTER 1

PASHTO: A SOCIOLINGUISTICAND DIALECT STUDY

1. FRAMEWORK AND PURPOSE

Data for the Pashto1 project was collected over a period ofabout two and a half years from 1988 through 1990. Pashto wordlists were collected from thirty-three scattered locations inPakistan and two locations in Afghanistan. The major focus ofthis study was Pakistan. In addition, a WaNeci word list wascollected from the Harnai area near Quetta as well as an OrmuRiword list from the Kaniguram area in South Waziristan. All ofthese lists were compared with each other in order to obtain a

lexical similarity count between each location. Recorded texttesting was also carried out between two locations, Peshawar andQuetta, in order to measure levels of comprehension betweenthese two areas. In addition, questionnaires/interview scheduleswere administered to nearly 140 participants from widelyscattered locations within Pashtoon territory. One of the mainpurposes of this study was to explore the questions of how manyPashto dialect groupings there really are today, and what the

criteria are for defining those groupings. A second, and equallyimportant, purpose for this study was to probe the areas of dialectopinion, language use, language attitudes, language vitality, andreported bilingualism within the Pashtoon community.

1

Generally the spellingPashto [pa}to] can be thought of as reflecting thesoftpronunciation of this language name. It is, however, a widely recognized pronunciation both in the literature and throughout Pashtoon territories. Thehard pronunciation might be reflected in the spelling Pakhto [paxto]. In thisdocument, the spelling Pashto is used throughout for uniformity of reference;when the spellingPashto is used it is not necessarily meant to refer to the soft

pronunciation.


20/194

Pashto, WaNeci, OrmuRi2

2. GENERAL BACKGROUND

2.1 Pashto Language Classification

Pashto has been classified as an Eastern Iranian languagewhich, according to MacKenzie, came out of the Aryan family oflanguages that divided into its distinct Indian and Iranian branches more than three millennia ago (1969:450). Griersondescribes what may have happened after this division:

The Eranian and the Indo-Aryan language eachdeveloped on its own line. In the earliest stages of their

separate growth they were very similar

We have divided the Eranian languages into twogroups, Persic and Non-Persic. The Non-Persicdialects are often classed together under the termMedic, a convenient, but inaccurate name. They werespoken in widely separate parts of Eran In thepresence of literary and official Persian, Medic, as aliterary language died a natural death, and survived onlyin local dialects of which we have no medival literaryrecords.

In course of time, these Medic dialects developedinto independent languages, some of which form thesubject of the present volume. These are the Ghalchahlanguages of the Pamirs, Pashto, OrmuRi, and Balochi all those dealt with in the present volume, are spoken

in the eastern part of the ancient Eran, they can, for the purposes of this Survey, be conveniently classedtogether under the name of the Eastern Group of theEranian languages (Grierson 1921:2-3).

2.2 The Geographic Boundaries of Pashtoon Territory

As might be expected, setting precise geographic boundarieswithin which Pashto is spoken is not an easy task. Some scholarshave, however, made an attempt to describe generally wherethese boundaries lie. Although a bit outdated in terms of naming


21/194

Pashto 3

political affiliations, the general description of the boundaries ofPashtoon territory given by Grierson seem basically true today:

Pashto is the language of the greater part ofAfghanistan. In the North-Western Frontier Provinceand the adjoining sphere of British influence, i.e. inwhat we may call British Afghanistan, it is spoken inthe Districts of Peshawar, Hazara, Bannu, Kohat, andDera Ismail Khan, and in the region between them andthe Afghan frontier. It is, moreover, the language of theindependent Yusufzai country which may conveniently

be called the Yaghistan, situated to the north of theseBritish districts, and including the countries of Swat,Buner, and Bajaur. The tract composed of these threecountries, the Afghan portion of Afghanistan, theBritish district above mentioned, and the Yaghistan is known as the Roh, that is to say, the Hill Country.The Roh is defined by the historian Firishta as thecountry extending, from north to south, from Swat and

Bajaur to Siwi and Bhakar, and from east to west, fromHasan Abdul to Kabul. It includes Kandahar.

In British territory the eastern boundary of Pashtomay be roughly taken as coinciding with the course ofthe Indus, although there are Pashto-speaking coloniesin the Hazara and Attock Districts on the Indian side ofthe river. After entering the district of Dera Ismail Khanthe eastern boundary gradually slopes away from the

Indus, leaving the lower parts of the valley in possession of Lahnda, and some thirty miles south ofthe town of Chaudhwan it meets Balochi. The southern boundary passes south of Quetta and throughShorawak, till it is stopped by the desert of Baluchistan.

Taking up the southern boundary of Pashtowhere we left it, after passing through Shorawak, in thedesert of Baluchistan, we find that it follows the eastern

and northern limits of that desert, with extensivecolonies down the rivers which run south through thewaste, to nearly the sixty-first degree of east longitude.


22/194


It then turns northwards up to about fifty miles south ofHerat, where it reaches its limit to the north-west. From

here the northern boundary runs nearly due east to theHazara country, in which tract the majority of theinhabitants do not employ Pashto but speak eitherPersian or a language of Mongolian origin. Skirting thewest, south, and east of the Hazara country, and justavoiding the town of Ghazni, it thence runs northwardsto the Hindukush. Thence leaving Laghman andKafiristan to its east and north, the boundary roughlyfollows the Kabul river down to Jalalabad, whence itruns up the Kunar so as to include the Yaghistan asalready explained. Speaking roughly, we may sum upthe above irregularly shaped block of Pashto-speakingterritory as including Southern and EasternAfghanistan, the country to the west of the Indus inBritish territory, from its southward bend to Dera IsmailKhan, and a strip of Northern Baluchistan (1921:5-6).

A more recent accounting by Penzl is much the same:

In Pakistan Pashto is spoken in the NorthwestFrontier Province in the districts of Peshawar, Hazara[Pakistan], Kohat, Bannu, Dera Ismail Khan, also in theterritories of Swat, Buner, and Bajaur. It is also spokenin northeastern Baluchistan in the Quetta-Pishin, theLoralai, Zhob, and Sibi areas; in the Punjab it is stillspoken in the border areas of Mianwali and Attock. Thewhole tribal area between Pakistan and Afghanistan isPashto-speaking.

In Afghanistan the Pashto-speaking area is in theEast, the South, and the Southwest. Pashto is spoken inthe entire Eastern [mashreqi] Province, which hasJelalabad [Dzhelaalaabaad] as its capital; in thesouthern [szhinubi] administrative province with

Gardez [Gardeez] as the principal city; in the southernand central parts of the province of Kabul outside of theHazara [Afghanistan] territory; in the entire province of


23/194

Pashto 5

Kandahar [Qandahaar]; and in most of the

administrative province of Farah (1955:1-2).

2.3 Population Estimates

Totally accurate population figures for Pashto are hard toobtain. Concerning the 1981 census of Pakistan, Kluck says:

As with all previous censuses and enumerations,these were estimates. Questions related to a mans

sisters, wives, mothers, or daughters are the subject ofgreat sensitivity. Even questions about the number ofwomen in a household are, in the Pakhtun view, aninvasion of privacy The tendency is for respondentsto overstate the number of men in a household andundercount the number of women (1984:85).

In spite of this uncertainty about exact figures, one can saywith confidence that there are vast numbers of speakers of Pashto

in both Pakistan and Afghanistan. Official 1981 total populationcensus figures for the Northwest Frontier Province (NWFP) were11,061,328. For Baluchistan the total population recorded was4,332,376 (The Europa World Year Book 1990, 2.1995). To geta better idea of what percentage of this total population might bePashto speaking, reference can be made to figures reported byKluck:

The 1981 census enumerates 10.9 million residents all but 20 or 30 percent Pakhtuns in the NWFP.There were, in addition, some 2.8 million officiallyregistered Afghan refugees who were overwhelminglyPakhtun (1984:85).

Comments by Penzl on the 1951 census estimate the totalpopulation of Pashto-speakers for both Pakistan and Afghanistanat around 13 million speakers, but again there is recognition of

possible inaccuracy in Penzls statement, No regular detailedcensus has been undertaken in Afghanistan. The officialestimates all appear high (1955:3).


24/194


With respect to the Pashto-speaking population inBaluchistan, no current figures were obtainable at the time of thiswriting. However, overall estimates of the total population of

Baluchistan in 1981 (4,332,376 people) indicate that the population is considerably smaller than that of the NWFP. Arough estimate may be somewhere between 1,000,000 and1,500,000 Baluchistan Pashto speakers. If this estimate is at allaccurate, it can be said that a relatively small percentage of thetotal Pashtoon population lives in Baluchistan.

2.4 Education in Pakistan

In general the state of education in Pakistan can best besummed up by a few comments taken from Kluck:

Attendance rates for the school-age populationremained low and drop-out rates high. The 1981 censuscounted less than one quarter of those over 10 years ofage as literate; the rate fell to 14 percent for women, 15

percent of the rural populace, and a scant 5.5 percent ofrural women

Despite substantial growth, female enrollmentremained low; in 1982-83 their portion of estimatedenrollment was some 28 percent, 30 percent of the primary students. Female enrollments were up from alow of 14 percent in 1947; in the 1960s and 1970s alonethey had gained 10 percentage points

A variety of problems have dogged the expansionof educational opportunities. Roughly half of primary-school-age children were enrolled in 1982-83. Rates ofabsenteeism, among students and teachers alike, werehigh. In the early 1980s half of all students dropped outbefore finishing the fifth year of their schooling. Drop-out rates for girls in primary school were extremelyhigh. Only two-thirds of girls entering first grade

continued on to second, only one in 10 continuedthrough the tenth grade (1984:125-6).

On a more positive note, however, this same source says:


25/194

Pashto 7

The number of schools grew dramatically, nearlytripling from 1960 through the early 1980s. Unlike theincrease in enrollments, the growth in schools was

spread relatively evenly among primary, middle, andadvanced schools (Kluck 1984:126).

One further point addressed by Kluck is:

The language of instruction, according to 1979 and1983 decrees, is Urdu. Implementation of instruction inUrdu has been hampered by a lack of adequately trainedteachers. Instruction in the private schools frequented

by the elite and aspiring middle class remains, de facto,English (1984:125).

One implication that can be drawn from this information isthat since the school setting is one of the primary domains whereUrdu is learned, and since a relatively small percentage of thetotal population of Pakistan has had opportunity to attend schoolto any great degree, there is a large percentage of that total population which has low proficiency in Urdu. This is most

poignant amongst the women who appear to have the leastopportunities. (This fact is supported by questionnaire responsesgathered from Pashtoon participants in the course of this presentstudy.) There is little doubt that the statements above describe thegeneral state of affairs amongst the Pashtoon community as well.

2.5 Comments on the Pashto Alphabet as Expressed inthe Literature

Unlike many of the smaller languages in Pakistan, Pashtohas had a fairly long literary tradition. Part of this tradition hasentailed the development of and subsequent discussion of issuessurrounding the Pashto alphabet and the symbols which representit. Penzl reports, The Arabic alphabet, usually in the Nashk

form, is universally used in Afghanistan for both Persian andPashto (1955:5). This would also appear to be true in Pakistan.


26/194


A number of authors have made comments about the theorigin of the Pashto alphabet speculating on how this writtenstandard, if a standard really exists, came to be. Penzl writes:

The correlation between the Kandahar phonemic pattern and the graphic pattern of the special Pashtosymbols of the Arabic alphabet is so close that we mustassume that these symbols were created in the area ofthe Kandahar dialect. Kandahar appears to be the cradleof the Pashto alphabet (1955:10).

In contrast to this point of view, Morgenstierne says:

When the orthography of Psht. was fixed in the16th century, the distinction between S7, Z7 and x, gseems still to have been preserved even among thenorth-eastern tribes, who were probably the creators ofPsht. literature (1932:17).

Concerning this controversy, MacKenzie says in his 1959article,A Standard Pashto:

It would be rash to decide this question onorthographic evidence alone, but there is this to be saidin support of the northeastern hypothesis. An earlierorthographic tradition than that now prevailing onceexisted (1959:233).

Whatever the actual case may be, it does appear that there isa standard written form of sorts in existence today which serves

to somewhat overshadow the differences that exist between someof the various dialects mentioned in the literature. This was alsothe opinion of scholars encountered in this present study whosaid that the writing system is the same all over, but the verbal pronunciation of words varies in different Pashtoon areas.MacKenzie writes:

The criteria of dialect differentiation in Pashto are

primarily phonological. With the use of an alphabetwhich disguises these phonological differences thelanguage has, therefore, been a literary vehicle, widelyunderstood, for at least four centuries. This literary


27/194

Pashto 9

language has long been referred to in the west as

common or standard Pashto without, seemingly anyreal attempt to define it (1959:231).

In his 1959 article entitled Standard Pashto and the Dialectsof Pashto, Penzl writes:

The standard Pashto orthography follows thephonemic distinctions as found in the Kandahar dialect.Even the speakers of dialects where the number [of]

phonemes differs from the Kandahar dialect use thisstandard orthography when they write. Even in theirdialect, e.g., as in Peshawar, zz [{] has coalesced withg, ss [s7] with kh, dz with z, ts with s, they accept theKandahar orthography as standard and try to make itsphonemic distinctions in writing (1959:12).

Current efforts to clarify the orthographic conventionsfurther support the idea that there is a standard writing system of

sorts. Even today the Pashtoon community as a whole, at least onthe academic level, is concerned about further developing awritten form of Pashto which is more standardized than in thepast. This is evident in the fact that in July of 1990 a symposiumon Pashto script was held by the Pashto Academy, University ofPeshawar; participants in this symposium sought to arrive at aconsensus on some points of controversy concerning differingforms of written Pashto. Scholars and interested people from all

segments of Pashtoon society and a wide range of Pashtoonlocations participated in the symposium.

3. DIALECTS OF PASHTO

3.1 Pashto Dialect Groupings as Expressed in ExistingLiterature

A number of sources discuss various dialect divisions withinthe Pashto language. One distinction which is almost universallymentioned in these sources is the distinction between hard andsoftPashto. Speakers of hard Pashto can be defined basically as


28/194


those who pronounce the letter as [x] while speakers of softPashto, in contrast, pronounce this same letter as [}]. On thistopic Grierson says:

Over the whole area in which it is spoken, thelanguage is essentially the same. This will to someextent be evident from the specimens which follow Such as they are they show that, while, as we go fromtribe to tribe there are slight differences in pronunciation and grammar, the specimens are allwritten in various forms of what is one and the samelanguage. Two main dialects are, however, recognised,

that of the north-east, and that of the south-west. Theymainly differ in pronunciation. The Afghans of theNorth-east pronounce the letter k4h4 and letter g,while those of the South-west pronounce them Sh4 andZh4, respectively (1921:7).

Another statement determines where Grierson thought thesetwo varieties to be spoken:

The North-Eastern dialect is spoken in the districtof Hazara, and over the greater part of the districts ofPeshawar and Kohat, but in the two latter the membersof the Khatak tribe use the South-Western dialect. Inthe districts of Bannu and Dera Ismail Khan the South-Western dialect is universal (1921:10).

In yet another statement, when speaking about South-Western Pashto speakers besides the Khataks, Grierson says:

Other speakers of the South-Western dialect arethe remaining Pathan tribes of Bannu, among whom theprincipal are Marwats, the Nyazais, the Bannuchis, andthe Wazirs (Grierson 1921:69).

Many other writers have also pointed out this major two- part division between Pashto varieties, but in later writings afiner distinction based on pronunciation is delineated. One such

writer is D.N. MacKenzie, who, in his 1959 article entitled AStandard Pashto, distinguishes four dialect areas based on fivedifferent phonemes. These are: South-west (Kandahar), South-


29/194

Pashto 11

east (Quetta), North-west (Central Ghilzai), and North-east

(Yusufzai). He summarizes the distinctions as presented in (1)(1959:232):2

(1)

SW SE NW NE(Kandahar) (Quetta) (Central (Yusufzai)

Ghilzai)

1. c [ts] c (s) (s)2. j [dz] j (z) (z)

3. { [] { { (j)

4. Z [] ({) G [j] (g)

5. S [ ] (}) x [] (x)

Other writers, Henderson (1983) and Skjrv (1989), alsodiscuss this basic four-part division of Pashto dialects. Penzl,who divides Afghanistan into the Eastern and Kandaharvarieties, describes a basic three-part division, leaving out anydistinction between the dialects of Kandahar and Quetta (1955:8-9).

Although a four-part distinction is helpful for marking outmajor divisions of Pashto, it is apparent from the literature that

this alone does not serve to capture all of the importantdistinctions. Perhaps the most clear case of this has to do with thevariety of Pashto that is spoken in Waziristan. This variety hasbeen singled out by some scholars as being very different fromthe varieties of Pashto spoken in the north. In reference to J.G.Lorimers 1902 work entitled Grammar and Vocabulary ofWaziri Pashto, Penzl says:

2 Symbols in brackets present an alternate phonetic writing system.MacKenzie has also placed some phonetic symbols in parentheses. He saysthese symbols reflect the fact that moving away from the south-westerndialect, there is a steady depletion in the inventory of consonant phonemes,owing to coincidence with existing phonemes (in parentheses) (1959:232).


30/194


J.G. Lorimer was prompted to write his Grammarand Vocabulary of Waziri Pashto (1902), because hefelt that nobody who spoke only the dialect of Peshawar

could make himself understood in Waziristan (1955:7).This is further underscored by information about the

Pashtoon tribes of Waziristan that J. G. Lorimer related toGrierson:

The dialects spoken by those tribes do not varygreatly from one another, but differ considerably inaccidence, vocabulary, and even idiom, from the

dialects spoken by the Pathans on the Kohat andPeshawar Frontiers, indeed an untraveled NorthernPathan and an untraveled Waziri meeting for the firsttime are scarcely intelligible to each other, and arecertain to misunderstand one another to some extent.Each, however, rapidly becomes able to understand theothers language, but I know of no instance of anorthern Pathan who has learned to speak Waziri

Pashto (J. G. Lorimer cited in Grierson 1921:96).The apparent uniqueness of Waziri Pashto, as described by

certain authors, can also be seen in some of the particulars of thegrammar and phonology. For example, Skjrv in his articleentitledPashto presents one conjugation of the verb to be andgives one single listing of verbal endings, both past and present,for what he callsstandardPashto, which looks to include at leastthe somewhat recognized standard Pashto varieties of Kandahar

and Peshawar (Yusufzai). For Waziri, however, he gives aseparate conjugation and list of verbal endings which whenexamined seems to differ considerably from that ofstandardPashto. Although this is but one small example of the uniquenature of the Waziri variety of Pashto, it may be an indicator ofthe overall uniqueness of this variety.

On the matter of Waziri phonology, several scholars pointout some of the particular unique vowel shifts that seem toexpress themselves most uniquely in this variety of Pashto.Morgenstierne says:


31/194

Pashto 13

In the east central dialects, from Afridi to Wazirithere is a tendency towards a change of the vowel

system. In those Waziri and Bannu dialects where thistendency is carried through most radically, we find o2, e2,2 for ordinary Psht. a2, o2, u2, and frequently a2 ore2 fora(1932:18).

In section 3.2 below, word list data collected for this presentstudy is presented. Lexical and phonological informationgathered and discussed in that section will serve as a furtherindicator of the uniqueness of Waziri.

In addition to the unique qualities found in Waziri, it alsoseems that other Pashto varieties exhibit qualities that are notspecifically revealed by the simple four-part division mentionedabove. Morgenstierne says:

the dialectal variety of Psht. [Pashto] is fargreater than that of Bal.[uchi]. And among the Afghans,the nomadic Ghilzais and the comparatively recent

invaders of Peshawar, Swat, etc. show the least amountof dialectal variation, while the central part of Pashtospeaking territory is the one which is most split up intodifferent dialects (1932:17).

Following this statement in his Report on a LinguisticMission to North-Western India, Morgenstierne then goes on tospell out specifically some of the peculiarities of the varieties ofcentral-area Pashto.

Likewise, concerning the varieties of Pashto spoken in the province of Baluchistan, earlier in the same volume,Morgenstierne says:

The most important Pathan tribes of the provinceare the Kakars and Spin (White), and Tor (Black)Tarins

Common to both Kakari and Tarin (and also tosome Ghilzai dialects such as Pur Khel) is the tendencytowards depalatalization of common Psht. (1932:11).

Despite these differences, however, Morgenstierne also says:


32/194


On the whole the differences existing betweenthese southernmost dialects of Psht. [Pashto] are notvery deepgoing, and do not affect essential parts of the

phonological or morphological system of the language.But still the Psht. dialects of Balochistan present severalarchaic or otherwise interesting forms, some of whichhave been mentioned in an article in the NTS [NorskTidsskrift for Sprogvidenskap] (1932:11-2).

3.2 Word List Findings

In this present study, comparison of Pashto word listscollected in thirty-four different locations indicates that there areat least three clearly identifiable groups of Pashto varieties inPakistan and perhaps a fourth less distinct grouping as well.These four will be described below. In addition, there is alanguage variety called WaNeci which is spoken in Baluchistan,near Quetta, which appears to be set off by itself because ofthe uniqueness of WaNeci, it will be dealt with separately inchapter 2. (See appendix B for a complete display of the wordlists collected from thirty-four locations.3)

Percentages of lexical similarity between each pair oflocations are displayed in figure (2). The method used forestablishing similarity, based on shared phonetic featuresbetween lexical items, is described in appendix A.1.

3 It should be noted that in the Pashto word lists and texts which appear in

appendices A and B, the symbols V and F do not represent distinctivephonemes. Also, no attempt was made to differentiate between an unstressed and eitherV or F. This position concerning Pashto is supported by Skjrvwho says, The opposition between /a/ and /F/ is neutralized in unstressed

position...other unstressed vowels as well may be realized as [F] (1989:388).


33/194

Pashto 15

P

eshawar

(2)Lexicalsimilaritym

atrixPashto

97Charsadda

96

97Mardan

92

9595

Swabi

95

9797

95

Madyan

94

9697

94

99

Mingora/Saidu

92

9695

94

99

98

Batagram

93

9595

93

98

97

100Baffa

88

9391

91

96

96

9999

Oghi

88

9091

88

94

94

9393

90

Dir

NorthernGroup

88

9092

90

93

93

9393

90

91

Bajaur

91

9394

93

95

94

9496

90

88

89

Moh

mand

90

9191

89

90

90

8988

86

87

87

89

Ningrahar

89

9291

89

92

90

9190

88

87

88

91

92

Shinwari

86

8988

85

89

89

8988

86

83

84

84

88

93

BarShilman

88

9088

84

88

88

8687

83

83

83

86

88

92

90

Mallagori

82

8482

80

84

83

8382

81

78

79

80

83

91

87

85

ZakhaKhelAfridi

91

9189

86

90

89

8987

86

85

85

88

90

95

91

92

92

JamrudAfridi

83

8683

83

84

83

8584

83

79

82

83

84

90

85

84

91

93

TirahAfridi

9610097

93

97

95

9694

92

91

90

93

91

92

88

88

84

91

85

Jallozai

93

9493

92

94

93

9393

89

87

90

90

88

90

88

87

84

89

85

95

Cherat

81

8281

78

81

79

8079

78

80

78

78

84

88

80

82

84

88

85

83

80

Parachinar

82

8383

80

83

82

8383

79

79

79

79

79

86

82

79

81

86

83

84

84

87

Hangu

MiddleSettled/TribalGroup

77

7878

74

79

77

7777

74

75

76

75

75

82

77

79

81

84

83

81

78

86

86

Thal

79

8180

76

81

80

8080

77

77

77

78

79

86

81

80

86

87

87

83

81

84

83

89

Karak

82

8585

80

85

84

8383

81

82

81

81

84

86

83

80

82

86

83

86

81

81

82

83

87LakkiMarwat

74

7775

73

78

76

7678

73

73

74

76

74

80

75

75

79

82

82

80

77

76

75

83

8883

Bannu

Ce

ntralGroup

81

8383

79

83

81

8281

78

80

78

81

82

86

80

81

82

86

84

83

82

84

80

83

8687

87

MiranShah

68

7171

68

72

70

7071

69

67

68

68

71

74

69

67

74

75

77

72

67

72

69

75

8281

80

87

Wana

80

8282

80

84

83

8483

82

78

79

79

84

83

80

78

77

80

76

84

80

78

75

70

7683

72

80

71

Quetta

78

8079

79

80

80

7978

78

75

78

77

83

81

77

76

76

80

75

81

78

76

74

69

7380

71

76

70

96

Chaman

SouthernGroup

81

8383

82

83

83

8282

80

77

81

80

86

85

79

79

79

82

77

84

82

79

75

72

7983

73

80

72

98

98

Pishin

79

7979

78

80

81

7979

78

75

77

77

83

80

78

77

75

77

75

80

78

76

71

67

7379

68

78

71

96

91

95Ka

kari(Pishin)

79

8180

81

81

81

8080

79

76

78

76

84

82

79

77

77

79

74

82

80

76

73

67

7479

71

77

70

96

97

9892

Kandahar

66

6969

67

72

71

6969

68

64

65

66

69

72

67

66

68

68

65

71

69

65

66

63

6772

65

70

64

75

71

7575

74

Waneci

27

3029

29

30

30

2928

28

27

25

27

27

31

29

25

30

31

30

30

29

30

29

30

3133

31

31

30

27

25

2725

28

27

Ormuri


34/194


3.2.1 Northern Group

The first group might be called the Northern group. (Othershave referred to it as Eastern or Northeastern Pashto.) This groupincludes the word list locations of Peshawar and Charsadda inDistrict Peshawar, Mardan and Swabi in District Mardan,Madyan and Mingora in District Swat, Batagram, Baffa, andOghi in District Mansehra, and Dir in District Dir. With only afew exceptions, all of the similarity counts between theselocations were 90 percent or above. In addition, within this larger Northern group there were sub-areas of greater similarity. Forexample, Madyan and Mingora, in District Swat, have 99 percentlexical similarity; Batagram, Baffa, and Oghi share 99 to 100 percent lexical similarity; and Peshawar and Charsadda are 97percent similar.

In contrast, similarity percentages between Northernlocations (including tribal locations) and nearly all of theSouthern-group localities (see 3.2.2 below) were in the 70s or

low 80s. Many percentages between the two major groups werein the 70s.

It should also be noted that although the locations of Cheratand Jallozai might be thought of as beingsoftvarieties of Pashto,and thus more akin to some of the southern varieties ofsoftPashto, word list counts clearly indicate that in most cases bothof these locations share greater than 90 percent lexical similarity

with all other Northern locations. In contrast, these two locationsgenerally have similarity percentages in the low 80s with softPashto varieties farther to the south (i.e., Waziristan, Bannu,Quetta, etc.) This would seem to make Cherat and Jallozai morea part of this Northern Pashto group than any other.

Word lists were also collected from nine northern tribes ornorthern tribal localities. These nine were: Bajaur, Mohmand,Shinwari, Bar Shilman, Mallagori, Zaxa Khel (Afridi), Jamrud

(Afridi), Tirah (Afridi), and one from Ningrahar (inAfghanistan). When comparing these nine word lists with wordlists taken from the settled area locations mentioned above, wordlist similarity counts were, for the most part, 88 percent or above.


35/194

Pashto 17

The two exceptions were Zaxa Khel Afridi and Tirah Afridi,which had similarity counts in the low 80 percent category withsettled area locations. From this, it might be concluded that the

northern tribal varieties of Pashto can basically be included in thelarger Northern Pashto group as well. It should be noted,however, that there are varying degrees of similarity within thislarger group, and some tribal varieties are a bit more divergent.

There is at least one sub-grouping amongst the triballocalities which should be mentioned. This sub-group might becalled the Afridi sub-group. Represented in this study are threeAfridi Pashto word lists: Zaxa Khel Afridi, Jamrud Afridi, andTirah Afridi. Similarity counts between these three were all 90percent or above, while similarity counts between these three andother tribal locations were generally at least a few percentagepoints lower.

3.2.2 Southern Group / Baluchistan Group

The second distinct grouping of Pashto locations might be

called the Baluchistan group. Others have referred to this asSouthern Pashto. Word lists were collected from four locations inBaluchistan as well as one from Kandahar in Afghanistan. Thefour Baluchistan locations were: Quetta, Pishin, Chaman, andPishin Kakari. All of these four have lexical similarity counts of90 percent or above. With the exception of the comparisonbetween Chaman and Pishin Kakari, all, in fact, have similaritycounts of 95 percent or above. Kandahar in Afghanistan also has

a high degree of similarity with the four Baluchistan locations,with similarity counts of 96 percent or above with all except thecomparison between Kandahar and Pishin Kakari, which was 92percent.

It should also be noted here that the similarity countsbetween the Baluchistan locations and all other locations were, inmost cases, below 80 percent. Looking specifically at Northernlocations, as compared with Baluchistan, about half of the

similarity percentages were below 80 percent while the other halfwere either 80 percent, or in the very low 80s. In particular, itshould be noted that no greater than 72 percent similarity was


36/194


calculated between any Baluchistan location and Wana (SouthWaziristan).

3.2.3 Central Group / Waziristan and Southern N.W.F.P.

A third distinct Pashto location is South Waziristan (Wanaword list collection site). Pashto is spoken quite differently therefrom almost any other place in Pashtoon territory (at least inPakistan). No Pashto word list has better than 77 percentsimilarity with Wana (South Waziristan) except for the localitiesof Miran Shah, Karak, Lakki Marwat, and Bannu, which are allin the same general area as South Waziristan.

Looking specifically at the comparison between Wana andall Peshawar-area and northern tribal locations, it is seen that thehighest percentage of similarity is 77 percent, between Wana andTirah Afridi. Most Northern locations have similarity counts withWana that are 70 percent or below. Also, as mentioned above,there is no greater than 72 percent similarity between Wana andany of the Baluchistan locations.

Although South Waziristan could be thought of as a Pashtovariety set apart by itself, based on relative similarity counts, itmight be grouped with North Waziristan (Miran Shah) andperhaps more loosely with other southern settled localities in theN.W.F.P. such as Karak4, Lakki Marwat, and Bannu. With onlyone exception (80 percent between Wana and Bannu), all ofthese five locations share greater than 80 percent lexicalsimilarity amongst themselves, whereas most similarity counts

between this five-location group and other locations are lower.However, if Wana is not considered a part of this lowertribal/Southern N.W.F.P. group, then with only one exception,5all other similarity counts are 86 percent or above.

4 It should be recognized that Karak groups rather loosely with thisCentral Pashto group since it also has a relatively high degree of shared

similarity with some word list data collection sites which do not fit well intothis group as a whole. For example, the similarity percentage between Karakand Thal is 89 percent. This is not surprising since Thal is relatively close toKarak District.

5 The similarity count between Lakki Marwat and Bannu was 83 percent.


37/194

Pashto 19

The conclusion that might be drawn about this CentralPashto group is that it is a much looser clustering than either the Northern group or the Baluchistan group. Generally there is a

higher degree of similarity amongst these five locations thanthere is between these five and other locations. This higherdegree of lexical similarity is, however, not as high as in theother two more distinct groupings previously mentioned (i.e.,Northern Group and Baluchistan Group). It is also important tonote that South Waziristan could be viewed as a variety of Pashtoapart from any other. It does share 87 percent similarity with North Waziristan, but with every other location in the Centralgroup of five, the similarity count is 80 to 82 percent. Manysimilarity counts between Wana and locations outside of theCentral group were 70 percent or below. These results supportthe distinctive nature of Waziri (especially that of SouthWaziristan) as was discussed in section 3.1 above.

3.2.4 Middle Settled Area / Middle Tribal Group

A fourth, less distinct, Pashto grouping can also beidentified based on word list comparisons. This group might becalled the Middle Tribal or Middle Settled Area group. Locationsincluded in this study were Hangu and Thal in District Kohat andParachinar in the Kurram Tribal Agency. All three locations havesimilarity counts of 85 percent or above amongst themselves,whereas the majority of lexical similarity counts between these

three and all other locations were 80 percent or below.Although some higher degree of similarity is shown

amongst these three locations, there is also a high degree ofsimilarity between some of them and certain northern triballocations. For example, the similarity percentages betweenParachinar and Jamrud Afridi and Parachinar and ShinwariPashto were both 88 percent. Other comparisons were generallynot as high, but this helps to demonstrate that this Middle Tribal

group is a much looser group than either the Northern group orthe Baluchistan group, if in fact it should be classified as a groupat all.


38/194


3.3 Comprehension of Recorded Speech

In addition to the comparison of word lists, some actualcomprehension testing was conducted by means of recorded texttests. (See appendix A.2 for a description of recorded text testmethodology.) Because of time constraints, research was limitedto testing between two sites, Peshawar and Quetta. The results ofthis testing shed some helpful light on just how well speakersfrom these two distant groups understand each other.

In this study two Peshawar stories were prepared for testing,the Electrocution story and the Auto Accident story. (Seeappendix C for a transcription and translation of these stories.)Both were hometown tested in Peshawar, and as a result, two orthree problem questions were removed from each text. The finalform of each test contained twelve questions. For each of the twostories the average extracted hometown score was nearly thesame. (The extracted score in this case was figured by removingthe problem questions and calculating the results based on the

remaining twelve questions.) For the Peshawar Electrocutionstory, the average extracted hometown score for twelve testsubjects was 98 percent with a standard deviation of 4. For theAuto Accident story the average extracted score was 97 percentwith a standard deviation of 6. The median in both cases was 100percent.

Both of the Peshawar stories were later taken to Quetta, in

Baluchistan. A hometown test was also prepared there andsubjects were given the hometown Quetta test for screening, afterwhich, they listened to and answered questions about eachPeshawar Pashto story.6 The fourteen Quetta subjects who weretested scored an average of 98 percent (with a standard deviationof 2) on their own hometown test, after which, each Quettasubject participated in comprehension testing on each of the twoPeshawar stories. In comparison with the scores achieved on the

6 Due primarily to lack of time, testing was not done in the reversedirection testing the Quetta story in Yusufzai/Peshawar territory. This issomething which probably should be done in the future to verify that Yusufzaireally is more widely understood than the Quetta dialect.


39/194

Pashto 21

hometown test, the scores were significantly lower (according toWilcoxon t test, p


40/194


understanding between the two groups. Both scores, however,indicate that most subjects understood much of each story.Further, many of those who did miss the particular questions

where vocabulary was a problem were still able to say the answeror some close sounding word, so filling in the particular lexicalslot did not seem to be a tremendous problem. This supports theidea that the main problem is vocabulary rather than syntax. Theidea that Pashto syntax and morphology is basically the samethroughout Pashtoon territory is supported by the scholarlyliterature (MacKenzie 1959, Penzl 1959, Jahani 1986).

One final note is that the two Peshawar stories were notalways played in the same order. Of the total fourteen subjectstested, seven of the subjects heard the Auto Accident story first,and seven heard the Electrocution story first. When analyzed, itwas found that there was no significant difference in how peoplescored based on the order of play.

3.4 Dialect Groupings Based on Native Speaker Opinion

In addition to word list and recorded text test data, interviewdata were collected regarding perceived dialect differences. Onthe interview schedule for this study there were a series of threequestions which asked how similar or different people thoughtother-area Pashto varieties were from their own. These threequestions were asked of over 130 subjects from fourteendifferent general areas in Pashtoon territory. The conclusionswhich can be drawn from this data help to confirm the

distinctions discussed above. A summary of the results is asfollows:

Responses to interview questions indicate that individualsfelt there is a general similarity between Pashto varieties in thesettled areas in the north. Pashto is spoken slightly differentlyfrom place to place (e.g., Swat, Peshawar, Hazara), but thedifferences do not appear to be very great. Within this generalcluster there were also perceived sub-clusters of greatersimilarity. Some of these were: Khalil and Mohmand; Mardan,Charsadda, Peshawar, and Buner; Swat and Hazara. Theoverarching perceived similarity between Northern Pashtovarieties corresponds nicely with the word list analysis presented


41/194

Pashto 23

above in which these locations were generally at least 90 percentsimilar to each other.

There was also a general consensus that the Pashto varieties

spoken in Baluchistan are similar. Quetta, Pishin, and Chaman,all in Baluchistan, were consistently said to be nearly the same by all twenty-six subjects from Baluchistan. Informalconversations with several Pashto speakers from bothBaluchistan and Kandahar, also support the idea that the Pashtoof Kandahar is very similar to that spoken in Quetta and thesurrounding area. This perceived similarity between SouthernPashto locations also corresponds nicely with the word list

analysis presented above in which these locations were all atleast 90 percent similar to each other.

In tribal territory, just adjacent to Peshawar, the AfridiPashto of Jamrud, Bara, Dara, and to a lesser extent, Zaxa Khel,were all perceived to be similar. This is supported by the wordlist analysis described above, where Jamrud Afridi and ZaxaKhel Afridi share 92 percent similarity. Also in tribal territorythe Pashto varieties of Shinwari, Shilmani, and Mallagori were

all perceived to be similar to each other. Again, these perceptionssupport the word list analysis where these three were all at least90 percent similar to each other.

Further to the south in District Kohat, Hangu, Thal, Doaba,and Ustarzai were perceived to be similar. And in District Bannu,Lakki (Marwat) was perceived to be quite similar to Kulachi andTank, which are both in D.I. Khan. In contrast, the Pashto ofBannu city area was perceived by many as being different frommost everything around it.7 Also, the Pashto of Wana (SouthWaziristan) and Mahsood (a particular tribe) were perceived tobe similar. Nine out of ten of the subjects from Wana said thatthe two were only a little different from one another.

7 The Pashto of Bannu was named as being quite different by 7 out of 8 ofthe Lakki and Kulachi subjects. It was also mentioned as being quite different

by 7 out of 10 subjects from South Waziristan. On the whole, Bannu wasnamed as being quite different by 70 out of 136 subjects from all over Pashtoonarea. Because the question was an open question asking where Pashto is spokenquite differently, those who did not mention Bannu may simply have notthought to mention it.


42/194


Also in support of the distinctions described above, mostindividuals from Northern Pashto areas named Waziristan,Bannu, Kohat, Quetta, Afridi, or Shinwari as varieties of Pashto

that are quite different from their own. Of these, Bannu,Waziristan, and Kohat were mentioned most often. Likewise,nine out of the ten subjects from South Waziristan mentionedPeshawar Pashto as being quite different from their own and sixmentioned Quetta. As noted above, seven of these ten also saidthat Bannu was quite different. These overall perceptions alsocorrespond nicely to the word list analysis described abovewhere, in general, localities which are more distant from oneanother, such as Peshawar, Waziristan, and Baluchistan, havelower shared lexical similarity percentages (generally in the 70sor low 80s) than do locations which are close to one another.

Within tribal territory, five out of six Afridi subjects fromJamrud mentioned Bannu as being quite different. Three alsoidentified Waziristan and four identified Shinwari as beingdifferent from their own. This last response is most interestingsince the Shinwari and Afridi tribes live relatively close to one

another. However, it should be noted that it may not always havebeen clear to each interviewee what exactly is meant by different.

In Southern Pashto territories, most subjects fromBaluchistan identified Peshawar Pashto as being quite differentfrom their own. A number also named Waziristan as being quitedifferent. A still smaller number of interviewees fromBaluchistan mentioned Zhob and Bannu Pashto as being

different from their own.

8

The fact that individuals perceivedPeshawar and Quetta Pashto to be different further supports theword list and comprehension analyses described above.

8 One further point to note about the responses by Baluchistan subjects isthat ten people identified Harnai as being a place where Pashto is spoken verydifferently. Harnai is located some 100 miles or so from Quetta and has beenidentified as the place where people speak a language/dialect that is a different

form of Pashto. The name often given to this variety is Waeci. Although thisgroup is reported to be relatively small in comparison to other groups in thearea, it would seem that many Pashto speakers from Quetta and the surroundingarea are aware of this language. Waeci will be dealt with in more detail inchapter 2 of this volume.


43/194

Pashto 25

Most interviewees were also asked how well they felt theyunderstood varieties of Pashto they thought were quite differentfrom their own. Of those 133 subjects, about half expressed the

idea that they could understand them or that they couldunderstand most. A slightly smaller number felt they couldunderstand a little or some. A few said they could understandsome varieties well. For those from Northern Pashto areas, themost often specifically mentioned varieties of Pashto named asbeing hard to understand were those of Waziristan and Bannu.

3.5 Pashto Dialect Groupings Summary and

Conclusions

In summary, current literature concerning dialects of Pashtosuggests that there are four dialects which can be differentiatedon the basis of the pronunciation of five phonemes. A more basicdistinction between hardand softPashto has also been pointedout by a number of authors. Although these distinctions arehelpful, it is clear that they are not adequate to describe all of theimportant differences which exist between the different varieties

of Pashto that are spoken today. This is especially clear in thecase of the type of Pashto spoken in the area of Waziristan.Existing literature as well as evidence collected in this presentstudy suggest that this variety of Pashto may be somewhatunique.

Although the literature which already exists concerningPashto dialects is helpful, this present study serves to make a

contribution toward clarifying where other important distinctionsexist between Pashto varieties. For example, when looking atMacKenzies four-part distinction, it is seen that the Pashto ofQuetta, Pakistan is separated from that spoken in Kandahar,Afghanistan. MacKenzie makes this distinction on the basis oftwo phonological distinctions, }/S and {/Z. Although thisinformation is accurate and useful, it can be seen from the wordlist data collected in this present study that Quetta and Kandaharshare 95 percent lexical similarity. This is a fairly high degree of

lexical similarity for two varieties of Pashto that MacKenzie andothers would put in separate dialect groups. Based on this highdegree of lexical similarity, it is perhaps more accurate to viewQuetta and Kandahar as members of the same group.


44/194


Another important clarification which is made in this studyis that the distinction between hardandsoftPashto is not alwaysthe most useful. Examination of existing literature along with the

word lists collected in this present study shows that the softvariety of Pashto is spoken over a wide area, all the way up intothe area of the Khattaks a relatively short distance fromPeshawar. According to MacKenzies four part distinction, thismeans that the Khattak variety of Pashto would be in theSoutheast group (Quetta). Yet when the word list locations ofJallozai and Cherat (in Khattak territory) are compared withword lists from other Pashto locations, it is seen that Jallozai andCherat are closer lexically to the varieties of Pashto spoken in thenorth (Peshawar area) than they are to any of the Southernvarieties. This makes sense since Jallozai and Cherat arerelatively close to Peshawar.

The distinction between hard and soft Pashto also lackssome description since there is considerable variation, especiallyin the softPashto group. As already mentioned, the locations ofJallozai and Cherat are lexically more similar to certain hard

varieties than to most othersoft varieties. Another example ofthis diversity is found in the variety of Pashto that is spoken inWaziristan and the surrounding area. According to thedistinctions made by MacKenzie and other researchers,Waziristan is a part of the softPashto group. However, as J. G.Lorimer points out and as the evidence in this present studyshows, the type of Pashto spoken in Waziristan and some of theadjoining areas is different from almost all other varieties,

including others which would also be considered to be softvarieties of Pashto, such as Quetta, Kandahar, etc. Under thesecircumstances it is perhaps better to put the variety of Pashtospoken in Waziristan and certain adjoining areas in a separatecategory. Thus, in this present study the category of CentralPashto has been used.

In short, the analysis which has been done in previousstudies is certainly valuable, but it is important that criteria otherthan phonological information be given more consideration whendrawing conclusions about dialect distinctions. Word list data,for example, take into account both phonological and lexicalinformation. Recorded text testing is designed to measure how


45/194

Pashto 27

well individuals can actually comprehend other languagevarieties. When these elements are added, a much fuller pictureof important dialect groupings can be seen.

4. BILINGUALISM / SECOND LANGUAGE

PROFICIENCY

4.1 Urdu

On the interview schedule used in this study there wereseveral questions asking subjects to give their opinion about their

own proficiency in Urdu and English as well as evaluating theproficiency of others in their own households and communities.Although asking someones opinion about bilingual ability isclearly not the best way to get a precise picture of the situation,these questions serve to give at least a general picture.

From the responses given, most subjects (all were male)claimed to have at least some ability in speaking Urdu. In fact, a

large majority claimed to be able to speak it well. Most of thesesubjects also claimed to be able to read Urdu. Of course it must be recognized that those who have gone to school would mostlikely have at least some minimal ability in speaking and readingUrdu. Since nearly all subjects questioned in this study had atleast some education, it is not surprising that they claim to haveat least some ability in Urdu.

As one might expect, interviewees felt that a higher

percentage of men in their community were able to speak Urduthan could women and children. Many subjects claimed that nowomen in their village could speak Urdu. In contrast, a higherpercentage of children were said to be able to speak Urdu, butstill many subjects said that either no children or an estimatedsmall percentage were actually able to speak it. Speculationsabout reading ability in Urdu basically fit the same pattern asestimates of speaking ability. Fewer women and children than

men were thought to be able to read Urdu.

It is not surprising that women in general are thought topossess low or no ability in Urdu since the place where Urdu is


46/194


most commonly learned is in school. Opportunities for women togo to school are fewer than for boys or men. Of those individualsasked, most either said that none of the women and girls in their

families attended school, or said that only the girls (presumablyyounger ones) would attend. In some more urban Pashtoon areasthe opportunities for girls seem to be greater than in rural areas, but still less than for boys. District Swat provides a goodexample where general enrollment and school figures providedby the District Education Office and the Girls Education Officein Swat show that at present there are far more boys schools andmale students than there are girls schools and female students.9In general, when subjects were asked to speculate on how manywomen in their village go to school, the majority either said nogirls or just a small percentage.

Aside from the school domain, interview responses indicatethat women do not travel much except out of necessity fordeath, marriage, or in case of illness. Whether or not they travelat all, their contacts are basically limited to extended familymembers who probably do not speak Urdu either. Contact with

Urdu speakers even among more traveled women is most likelylimited.

It should also be noted that reported language use indicatesthat Pashto is the exclusive language used in four out of sixdomains specifically mentioned in the interviews (i.e., home,mosque, jirga, and speaking to women). As for the remainingtwo domains, some subjects did claim to use Urdu as well as

Pashto in the bazaar, and of course those who go to school areexpected in most cases to operate at least partially in Urdu. In thehome domain, every subject claimed to speak Pashto exclusively.

9 Figures for District Swat which were obtained in late 1990 stated thatthere are 266 girls primary schools, 11 girls middle schools, and 7 girls highschools. In contrast, there were reported to be 718 boys primary schools, 75

boys middle schools, and 90 boys high schools. In the same way, with regard tototal numbers of students, there were reported to be 27,894 girls in primary and

middle school in Swat. Figures for girls in high school were not available. Bycomparison, there were reported to be 159,674 boys in primary through highschool. For primary and middle school alone there were reported to be 129,410male students, almost five times as many as the number reported for femaleenrolment.


47/194

Pashto 29

Since women spend most of their time in the home, this probablyoffers little opportunity for them to actually learn Urdu.

Another potential source for learning Urdu is throughtelevision and radio programs. Although it is hard to determinehow much women in the home actually listen to the radio orwatch television, it should be noted that a large majority ofsubjects who were asked claimed to have a television in theirhome. Fewer of those in outlying areas said they had a televisionin their home. When subjects were asked how much they watchPashto television programs, only a small percentage said they donot watch them. It is almost sure that those who do watch Pashtoprograms watch Urdu and other language programs as well, thusproviding an opportunity for women and those who travel less to be exposed to Urdu and maybe passively acquire someunderstanding of it.

The same can also be said about radio programs. Whenasked whether they listen to Pashto radio programs,

sociolinguistic survey of northern pakistan volume 4

Documents